Syntheses and DNA photocleavage by mono- and bis-phenothiazinium-piperazinexylene intercalators

Article abstract of DOI:10.1016/j.tet.2008.01.105

Chromophore systems consisting of one (compound 5) or two (compound 6) phenothiazine rings covalently attached to a bis-piperazinexylene chain were synthesized and evaluated as DNA photocleaving agents. In the presence of DNA, the compounds were shown to

Full text of DOI:10.1016/j.tet.2008.01.105

Available online at www.sciencedirect.com
Tetrahedron 64 (2008) 3429e3436
www.elsevier.com/locate/tet
Syntheses and DNA photocleavage by mono- and
bis-phenothiaziniumepiperazinexylene intercalators
a
Beth Wilson , Marıa-Jose Fernandez , Antonio Lorente , Kathryn B. Grant
b
b,
a,
*
*
´
´
´
a Department of Chemistry, Georgia State University, PO Box 4098, Atlanta, GA 30302-4098, USA
b Departamento de Qu´ımica Orga´nica, Universidad de Alcala´, 28871-Alcala´ de Henares, Madrid, Spain
Received 10 October 2007; received in revised form 22 January 2008; accepted 24 January 2008
Available online 31 January 2008
Abstract
Chromophore systems consisting of one (compound 5) or two (compound 6) phenothiazine rings covalently attached to a bis-piperazine-
xylene chain were synthesized and evaluated as DNA photocleaving agents. In the presence of DNA, the compounds were shown to monointer-
calate in their deaggregated forms and to strongly absorb red wavelengths of light. Reactions containing micromolar concentrations of
compound produced robust photocleavage of plasmid DNA under near-physiological conditions of temperature and pH (22 ꢀC and pH 7.0).
Phenothiazines 5 and 6 increased the Tm of calf thymus DNA by 17 and 19 ꢀC, indicating that significant levels of duplex stabilization were
produced.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: DNA strand breaks; Light; Phenothiazine; Piperazine; Xylene
1. Introduction
include strong absorption5 within the therapeutic window
for PDT (600e800 nm),9 low cellular toxicity in the absence
of light,5 and selective accumulation of phenothiazines in
tumor cells.5,6 Accordingly, the phenothiazinium compounds
methylene blue (lmax¼665 nm, 3¼8.16ꢁ104 Mꢂ1 cmꢂ1)4 and
toluidine blue (lmax¼636 nm, 3¼7.45ꢁ104 Mꢂ1 cmꢂ1)4 have
been used in a clinical setting to produce efficient photoactivity
against AIDS-related Kaposi’s sarcoma.10 Moreover, phenothi-
azines have exhibited efficient in vitro and in vivo photoinacti-
Phenothiazines are redox active, basic dyes that have been
utilized as electrophores in supramolecular assemblies1 and in
photogalvanic cells for potential solar energy conversion.2 In
biological systems, phenothiazines have shown a preference
for interacting with GC base pairs within double-helical
DNA, where they bind by intercalation.3,4 Because light-sensi-
tized phenothiazines are efficient generators of singlet oxygen
(1O2),5 they produce extensive photooxidative damage to
DNA4,6 by forming alkaline labile lesions and direct strand
breaks at guanine bases.7,8 In this regard, phenothiazines have
been attracting considerable attention for use as photosensitizers
in photodynamic therapy (PDT), a new therapeutic approach in
which red light is used to treat malignant tumors and non-
cancerous diseases.5,6 The advantages of phenothiazines
vation of bacterial, fungal, and viral pathogens.4e6,11e13
A
number of these studies have directly linked light-sensitized
DNA cleavage to the photodynamic action of methylene
blue.4,5,11,12
Herein we describe the syntheses and characterization of
two phenothiazinium photonucleases containing either one or
two phenothiazine rings covalently attached to a bis-piperazine-
xylene linker (7-dimethylamino-3-(1,10-[1,4-phenylenebis-
(methylene)bispiperazine])phenothiazin-5-ium iodide (5) and
N,N0-bis[(7-dimethylamino)phenothiazin-5-ium-3-yl]-1,10-[1,4-
phenylenebis(methylene)bispiperazine] diiodide (6)). In addi-
tion to evaluating DNA photocleavage efficiencies, we have
* Corresponding authors. Tel.: þ34 91 885 4691; fax: þ34 91 885 46 86
(A.L.); tel.: þ1 404 413 5522; fax: þ1 404 413 5505 (K.B.G.).
E-mail addresses: antonio.lorente@uah.es (A. Lorente), kbgrant@gsu.edu
(K.B. Grant).
0040-4020/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.01.105

3430
B. Wilson et al. / Tetrahedron 64 (2008) 3429e3436
compared binding affinities, and have determined the DNA bind-
ing modes of the two compounds.
for 24 h (Scheme 2). (Under acidic conditions, we obtained
mixtures of starting material with minor amounts of product
even after 48 h of reflux.)
In our first attempts to synthesize mono-phenothiazine 5,
the molar equivalents of 2 and 4 were varied and alternative
solvents and reaction times were employed (Table 1). Ulti-
mately, we found that treating 1 mol equiv of each reagent in
methanol afforded compound 5 in 34% yield (Scheme 3, Table
1). A small quantity of impure compound 6 (10%) was also re-
covered. (We had anticipated that the latter transformation
might require more effort, since two aromatic nucleophilic ad-
ditions (compound 6) versus one aromatic nucleophilic addi-
tion (compound 5) would be needed.) After a number of
procedural modifications, the synthesis of compound 6 was
achieved in 41% yield by reacting 2 mol equiv of phenothi-
azine 2 with 1 mol equiv of the bis-piperazinexylene linker 4
in the presence of a base (cesium carbonate) and DMF as sol-
vent (Scheme 3). Although several syntheses of 3,7-disubsti-
tuted phenothiazin-5-ium salts have appeared in the
literature,18,20,21 to the best of our knowledge, no synthetic ac-
counts of bis-phenothiazinium salts incorporating our design
features have been reported.
2. Results and discussion
2.1. Synthesis
We incorporated two key structural features into the design
of 5 and 6. A phenothiazine based on methylene blue was se-
lected as the chromophore, due to its ability to intercalate into
DNA3,4 and effect efficient DNA photooxidation within the
PDT therapeutic window (600e800 nm).14 Bis-piperazine-
xylene (4) was employed as a linker to enhance duplex stability
and DNA binding affinity. In intercalating and groove binding
compounds, xylene15 and piperazine16,17 rings have been used
to stabilize DNA by making extensive van der Waals contacts
within the nucleic acid grooves. In the case of piperazine, hy-
drogen bonding and electrostatic interactions are also possi-
ble.16,17 Finally, the structural composition of compound 5
(one intercalating phenothiazine ring and a terminal piper-
azino groove binding element) versus that of compound 6
(two intercalating ring systems) was intentionally included
in our design rationale in order to evaluate DNA photocleav-
age and binding affinity as function of DNA binding mode.
Known compounds phenothiazin-5-ium tetraiodide hydrate
(1),18 3-(dimethylamino)phenothiazin-5-ium triiodide (2)18
(Scheme 1), and 1,10-[1,4-phenylenebis(methylene)]bis(4-pi-
perazinecarbaldehyde) (3)19 (Scheme 2) were prepared ac-
cording to published procedures. (In the case of compound
2, the reaction proceeded in higher yield when chloroform
was employed as solvent in lieu of methanol.) Boykin et al.
utilized HCl to hydrolyze compound 3.19 However, we found
that bis-piperazinexylene 4 was more readily obtained by ba-
sic hydrolysis. Thus, known compound 419 was prepared in
78% yield by refluxing 3 in aqueous potassium hydroxide
Table 1
Synthesis of compound 5
Trial Compound Compound Solvent
2 (mmol)
Timea (h) Yield (%)
4 (mmol)
Type
mL
1
3
2
4
0.161
0.076
0.164
0.815
0.322
0.153
0.164
0.815
CHCl3
CHCl3
12
7
24
72
24
72
Trace
16
15
CH3OH 10
CH3OH 25
34
a
All reactions were run at 22 ꢀC over the time intervals indicated.
2.2. UVevisible spectra
Phenothiazines undergo dimerization in aqueous solutions as
a function ofincreased ionic strength and phenothiazine concen-
tration. In general, the absorption band of the dimer is hypochro-
mic and blue-shifted with respect to the absorption band of
corresponding monomer.4,6,22,23 Taking this into account,
H
N
S
N
S
N
b
a
S
N(CH3)2
.
1.5 I2 H2O
I
I3
1
2
Scheme 1. Preparation of 1 and 2. Reagents and conditions: (a) I2, CHCl3,
5 ꢀC, 2 h, 80%; (b) HN(CH3)2 in CH3OH, rt, 4 h, 55%.
+
2
4
a
b
1
4
9
6
O
N
S
2
N
S
8
α
H
N
N
β
H
N
N(CH3)2
N
N
N(CH3)2
Br
Br
N
N
N
N
O
N
N
N
N
b
a
H
N
I
2I
S
N
H
N
N
H
N(CH3)2
N
H
β '
H
α'
O
N
5
3
4
6
Scheme 2. Preparation of 3 and 4. Reagents and conditions: (a) K2CO3,
CH3OH, reflux, 24 h, 75%; (b) KOH, 3:2 CH3CH2OHeH2O (v/v), reflux,
24 h, 78%.
Scheme 3. Syntheses of compounds 5 and 6. Reagents and conditions: (a)
CH3OH, rt, 72 h, 34%; (b) Cs2CO3, DMF, rt, 48 h, 41%.

B. Wilson et al. / Tetrahedron 64 (2008) 3429e3436
3431
Table 2
Absorbance dataa
UVevisible spectra of 10 mM concentrations of compounds 5
and 6 were recorded in the absence and presence of 1% (w/v) so-
dium dodecyl sulfate (SDS), a negatively charged surfactant that
disrupts phenothiazine dimerization (Fig. 1, Table 2). Upon the
addition of SDS, it is evident that a pronounced red shift and sig-
nificant hyperchromicity are produced only in the case of bis-
phenothiazine 6, indicating that this compound forms dimers
under the experimental conditions employed.
Compound
lmax (nm)
3ꢁ104 (Mꢂ1 cmꢂ1
4.56
5.81
na
)
3ꢁ104 (Mꢂ1 (bp) cmꢂ1
na
na
)
5
662
660
665
580
662
665
5þ1% SDS
5þCT DNA
6
3.66
na
nd
6þ1% SDS
6þCT DNA
a
14.6
na
na
7.80
Extinction coefficients for compounds 5 and 6 were determined in 10 mM
sodium phosphate buffer pH 7.0 using solutions containing 1e10 mM of dye in
the absence and presence of 1% SDS (w/v) or 38e380 mM bp CT DNA. The
samples containing DNA were pre-equilibrated for 12 h in the dark at 22 ꢀC;
na¼not applicable; nd¼not determined due to aggregation in buffer.
1.50
A
Compound 5
1.25
concentrations of compounds 5 and 6 were recorded in the pres-
ence of 380 mM bp of calf thymus (CT) DNA (Fig. 1, Table 2).
The data indicate that compounds 5 and 6 exhibit bathochromic-
ity and hypochromicity when SDS is replaced by nucleic acid.
Upon the addition of 1.0% SDS, the lmax values of 10 mM
1.00
0.75
0.50
0.25
0.00
of compounds 5 and 6 are 660 nm (3¼5.81ꢁ104 Mꢂ1 cmꢂ1
)
and 662 nm (3¼14.6ꢁ104 Mꢂ1 cmꢂ1), respectively. In the
presence of CT DNA, the respective lmax values are 665 nm
(3¼3.66ꢁ104 Mꢂ1 (bp) cmꢂ1
)
and
665 nm
(3¼7.80ꢁ
104 Mꢂ1 (bp) cmꢂ1). It can be inferred from these data that
both compounds are in their monomeric, deaggregated forms
and are likely to be intercalated within the DNA. More impor-
tantly, DNA-bound phenothiazines 5 and 6 exhibit absorbance
maxima well within the therapeutic window used in photody-
namic therapy (600e800 nm).
400
450
500
550
600
650
Wavelength (nm)
700
750
800
1.50
B
Compound 6
2.3. Viscometric measurements
1.25
1.00
0.75
0.50
0.25
0.00
Our next goal was to obtain direct evidence of DNA interca-
lation. Viscosity measurements provide a simple and stringent
assay for determining the binding modes of DNA interacting
ligands. In order for an intercalator to insert itself between
adjacent base pairs within duplex DNA, the helix must first
unwind to create a gap large enough to accommodate the incom-
ing ligand. The unwinding process increases the contour length
of the helix, and as a result, the DNA becomes more viscous.
Because groove binding compounds do not require a base pair
gap, DNA lengthening does not occur, and viscosity is not
significantly changed. Thus, viscometric measurements can
be employed to distinguish between different DNA binding
modes.24 According to the theory of Cohen and Eisenberg, lin-
ear plots of the cubed root of the relative DNAviscosity (h/h0)1/3
versus the molar ratio of bound ligand to DNA bp (r) should
result in a slope close to 0.0 for groove binding compounds
and 1.0for classical monointercalators.25 Theslope of abisinter-
calator is then expected to be approximately twice that observed
in the case of monointercalation.
400
450
500
550
600
650
Wavelength (nm)
700
750
800
Figure 1. UVevisible spectra recorded at 22 ꢀC in 10 mM sodium phosphate
buffer pH 7.0 of: (A) 10 mM of compound 5 (line with squares) in the presence
of 380 mM bp CT DNA (solid line) or 1% SDS (w/v) (dashed line); (B) 10 mM
of compound 6 (line with squares) in the presence of 380 mM bp CT DNA
(solid line) or 1% SDS (w/v) (dashed line). The samples containing DNA
were pre-equilibrated for 12 h in the dark at 22 ꢀC.
Viscometric measurements were carried out using individ-
ual, pre-equilibrated solutions consisting of 200 mM bp of
CT DNA and 10 mM sodium phosphate buffer pH 7.0 in the
presence (h) and the absence (h0) of dye. The viscosity data
were plotted as (h/h0)1/3 versus r, for r values ranging from
0.01 to 0.06 (Fig. 2). The slopes obtained from the mono-
DNA base-stacking interactions give rise to bathochromic
wavelength shifts and hypochromic absorption in the electronic
spectra of the majority of DNA intercalators. Thus, as a prelimi-
nary test for DNA intercalation, UVevisible spectra of 10 mM

3432
B. Wilson et al. / Tetrahedron 64 (2008) 3429e3436
1.10
1.08
1.06
1.04
1.02
1.00
0.980
A
Lane
1
2
3
4
5
6
7
8
N
S
[dye μM]
0
+
10
-
10
+
5
2
1
+
0.5 0.25
light
+
+
+
+
% nicked 17 13
96
89
59
45
34
27
B
Lane
9
10
11
12
13
14
15
16
N
S
[dye μM]
0
10
-
10
+
5
+
2
+
1
+
0.5 0.25
0
0.01
0.02
0.03
r = [dye]/[DNA bp]
0.04
0.05
0.06
0.07
light
+
+
+
% nicked 12
11
93
79
59
64
34
28
Figure 2. Viscometric measurements of 200 mM bp calf thymus DNA pre-
equilibrated with 0, 2, 4, 6, 8, 10, 12 mM of compound 5 or compound 6 for
12 h at 22 ꢀC (10 mM sodium phosphate buffer pH 7.0): compound 5 (,,
slope¼0.99, R¼0.972) and compound 6 (C, slope¼1.20, R¼0.990).
Figure 3. (A) Compound 5 and (B) compound 6: 1% nondenaturing agarose
gels showing photocleavage of pUC19 plasmid DNA. Samples contained
10 mM sodium phosphate buffer pH 7.0 and 38 mM bp DNA in the absence
and presence of dye. After being equilibrated for 12 h in the dark at 22 ꢀC,
the samples were aerobically irradiated for 60 min at 22 ꢀC in a Rayonet Pho-
tochemical reactor fitted with twelve 575 nm lamps. Lanes 1 and 9: DNA con-
trols (no dye). Lanes 3e8: 10e0.25 mM compound 5. Lanes 11e16: 10e
0.25 mM compound 6. Lanes 2 and 10: 10 mM 5 and 10 mM 6 (no hn). Abbre-
viations: N¼nicked; S¼supercoiled.
and bis-phenothiazine plots were 0.99 and 1.20, respectively.
Clearly, compounds 5 and 6 both interact with CT DNA as
monofunctional intercalators.
The neighbor-exclusion principle26 states that every second
base pair site along the length of the DNA double-helix will
remain unoccupied during intercalation. In order to prevent
violation of nearest neighbor-exclusion, bisintercalators
Lane 11) for samples irradiated in the presence 10 mM of 5
and 10 mM of 6, respectively. In addition, both phenothiazines
exhibited detectable levels of photocleavage at concentrations
of compound as low as 0.25 mM (Lanes 8 and 16). Cleavage
was minimal in parallel control reactions run in the dark (Lanes
2 and 10) and in samples irradiated for 60 min in the absence of
phenothiazine (Lanes 1 and 9).
We then employed 10 mM concentrations of compounds 5
and 6 to monitor reaction kinetics. Individual samples contain-
ing 38 mM bp pUC19 plasmid DNA in 10 mM sodium phos-
phate buffer pH 7.0 were irradiated as described above, at
time points ranging from 5 to 60 min. As shown in Figure 4,
DNA photocleavage by compounds 5 and 6 was substantial af-
ter only 5 min of irradiation and steadily increased as a function
of increasing irradiation time. Maximal levels of nicked plas-
mid (93 and 95%) were generated at the 60 min time point.
should possess a linker chain whose length is ꢃ10 A.27 While
˚
the bis-piperazinexylene linker of compound 6 meets this re-
quirement,28 it is conceivable that conformation constraints
imposed by DNA lower the free energy for intercalative bind-
ing of a second ring, preventing its insertion into the nucleic
acid duplex.29 An additional consideration is linker rigidity.
Partial double bond character between the C3 atom of pheno-
thiazine and the proximal nitrogen atom of piperazine may
result if there is significant delocalization of the nitrogen
lone-pair electrons. Such a restriction might preclude interca-
lation of the second phenothiazine ring of compound 6.
2.4. DNA photocleavage
In our next experiment, DNA photocleavage was evaluated
as a function of decreasing dye concentration. Individual reac-
tions consisted of 10e0.25 mM of compound 5 or 6 pre-equili-
brated with 38 mM bp pUC19 plasmid DNA in 10 mM sodium
phosphate buffer pH 7.0. The samples were aerobically irradi-
ated for 60 min in a ventilated Rayonet Photochemical Reactor
fitted with twelve 575 nm lamps (spectral output 400e650 nm).
DNA direct strand breaks were then detected on a 1.0% nonde-
naturing agarose gel stained with ethidium bromide. As shown
in Figure 3, compounds 5 and 6 produced robust levels of DNA
photocleavage in a concentration dependent manner. The high-
est DNA cleavage yields (conversion of supercoiled plasmid to
the nicked form) were 96% (Fig. 3, Lane 3) and 93% (Fig. 3,
2.5. DNA thermal denaturation
During intercalation and/or groove binding, free energy
contributions arising from pep, van der Waals, electrostatic,
and hydrogen bonding interactions between ligands and DNA
stabilize the DNA double-helix. As a result, the melting tem-
perature (Tm) of the helix is increased. Because the Tm value of
ligand-bound DNA is related to the affinity of the ligand for
the nucleic acid, melting assays have been used to access
the relative binding affinities of compounds that associate
with double-helical DNA.30

B. Wilson et al. / Tetrahedron 64 (2008) 3429e3436
3433
for adriamycin, and 14.7 ꢀC for quinacrine.31 Clearly, the DTm
values produced by 5 and 6 are significant, providing evidence
for strong binding of both compounds to DNA and consider-
able levels of duplex stabilization.
100
80
Because the DTm values produced by compounds 5 and 6
are nearly equivalent, it can be inferred that the DNA binding
affinities of both dyes are similar. This result is consistent with
our observation that 5 and 6 interact with CT DNA as mono-
functional intercalators. It is also consistent with the dicationic
nature of the dyes at pH 7.0, the experimental conditions
employed here. Typically, basic phenothiazines have pKa
values greater than 12.32 In the case of compound 5, the bis-
piperazinexylene substituent would be expected to bear one
positive charge (the secondary amino group of the terminal
piperazine), as the piperazine tertiary amino groups would
not be protonated due a reduction in charge density arising
from the electron withdrawing effects of the xylene19 and phe-
nothiazine rings.32 Therefore, the dicationic charge density of
compound 6 is imparted by the two phenothiazines, while in
compound 5 the bis-piperazinexylene substituent is expected
to furnish one positive charge and the phenothiazine ring
the other positive charge. The melting temperature and viscos-
ity data can also be used to interpret the results of the absor-
bance and photocleavage experiments presented in this
paper. Although bis-phenothiazine 6 possesses two DNA-pho-
tosensitizing phenothiazinium rings and absorbs light more
strongly than mono-phenothiazine 5 in the DNA-bound form
(Fig. 1, Table 2), 5 and 6 both produce equivalent levels of
photocleavage (Figs. 3 and 4). Furthermore, both compounds
monointercalate and are likely to possess similar DNA binding
affinities. Taken together, these findings suggest that, similar to
methylene blue,7 intercalative binding promotes efficient pho-
tosensitization of direct DNA strand breaks by the phenothia-
zine chromophores in 5 and 6. In the case of methylene blue,
van der Putten and co-workers observed a significant reduction
in O2-dependent DNA photocleavage in the presence of
20 mM of Mg2þ, a magnesium ion concentration which
changed binding of the dye from an intercalative to a Mg2þ-in-
sensitive external binding mode.7 Levels of photocleavage
were further reduced when the singlet oxygen scavenger so-
dium azide was added to this reaction. The authors concluded
that methylene blue must be in close proximity to reactive sites
within the DNA in order for efficient O2-dependent photo-
cleavage to occur.
60
40
20
0
0
5
10
20
30
Time (min)
40
50
60
Figure 4. % DNA photocleavage of 38 mM bp pUC19 plasmid DNA in the
presence of 10 mM compound 5 (white bars) and 10 mM compound 6 (gray
bars) (10 mM sodium phosphate buffer pH 7.0 at 22 ꢀC). After pre-equilibra-
tion for 12 h in the dark at 22 ꢀC, individual samples were aerobically irradi-
ated for 0, 5, 10, 20, 30, 40, 50, and 60 min at 22 ꢀC in a Rayonet
Photochemical reactor fitted with twelve 575 nm lamps. The black bar repre-
sents DNA that was irradiated for 60 min in the absence of dye (10 mM
sodium phosphate buffer pH 7.0).
Taking into account the neighbor-exclusion principle,26 we
recorded DNA melting isotherms at a dye to DNA bp molar
ratio of 0.6 (Fig. 5). Compared to CT DNA (Tm¼69 ꢀC), com-
pound 5 exhibited a Tm of 86 ꢀC (DTm¼17 ꢀC), while com-
pound 6 produced a Tm of 88 ꢀC (DTm¼19 ꢀC). Some
typical DTm values for positively charged, DNA monointerca-
lators similar in design to 5 and 6 (i.e., positively charged in-
tercalators consisting of three fused aromatic rings attached to
a groove binding side chain) are: 7.2 ꢀC for ethidium, 13.1 ꢀC
1.00
0.80
0.60
0.40
0.20
0.00
3. Conclusions
In summary, we report the syntheses of cationic photonu-
clease 5 and 6, consisting of either one (5) or two (6) pheno-
thiazine rings covalently attached to a bis-piperazinexylene
linker. When bound to DNA, the phenothiazines generate
significant levels of duplex stabilization and exhibit strong
absorbance well within the 600e800 nm therapeutic window
required for photodynamic cancer therapy. At micromolar
concentrations of compound, robust levels of DNA photo-
cleavage are produced under near-physiological conditions
of temperature and pH (22 ꢀC and pH 7.0). Thus, the systems
30
40
50
60
70
Temperature (°C)
80
90
100
Figure 5. Melting isotherms of: 12.5 mM bp CT DNA in 10 mM sodium phos-
phate buffer pH 7.0: CT DNA ( , T ¼69 ꢀC); CT DNA with 7.5 mM of com-
6
m
pound 5 (,, Tm¼86 ꢀC); CT DNA with 7.5 mM of compound 6 (C,
Tm¼88 ꢀC).

3434
B. Wilson et al. / Tetrahedron 64 (2008) 3429e3436
described in the report may serve as a good starting point for
the development of new phototherapeutic agents.
acetone-d6, d (ppm)): 8.01 (m, 2H), 7.92 (m, 2H), 7.64 (m,
4H). 13C NMR (75 MHz, acetone-d6, d (ppm)): 153.6, 130.7,
129.5, 128.6, 125.5, 123.5. IR (film, n (cmꢂ1)): 2967, 1558,
1467, 1440, 1311, 1233, 1131, 1067, 1023, 841, 705. LRMS
(ESI): m/z calcd for C12H8NS [M]þ 198.04, found 199.0.
4. Experimental
4.1. General
4.2.2. 3-(Dimethylamino)phenothiazin-5-ium triiodide (2)
To a solution of phenothiazin-5-ium tetraiodide hydrate
(0.400 g, 0.553 mmol) in 20 mL of chloroform was added
dimethylamine in methanol (0.553 mL, 1.11 mmol) dropwise
over 4 h. Reaction progress was monitored by silica gel
TLC. The resulting precipitate was filtered, washed with chlo-
roform, and allowed to air dry. Product 2 (189 mg, 55%) was
obtained as a dark-blue solid. TLC (3:7 10% aqueous ammo-
Melting points were determined using a Stuart Scientific
1
model SMP10 apparatus. H and 13C NMR spectra were re-
corded at 300 and 75 MHz, respectively, on either a Varian
Unity One or a Varian Mercury-VX-300 spectrometer. Carbon
and proton assignments were based on HSQC and HMBC
experiments. Infrared spectra were acquired with an FT-IR
PerkineElmer Spectrum One spectrophotometer. Elemental
analyses (CHNS) were done on a Leco CHNS-932 automatic
analyzer. Iodine composition was performed by oxygen flask
combustion and by ion chromatography (Atlantic Microlabs,
Inc. Norcross, GA). Electrospray ionization (ESI) mass spectra
were generated using a Micromass Q-Tof hybrid mass spec-
trometer. UVevisible spectra were recorded on a UV-1601
Shimadzu spectrophotometer and thermal melting curves on
a Cary Bio 100 UVevis spectrophotometer. Merck silica gel
60 (230e400 ASTM mesh) was employed for flash column
chromatography.
1
nium acetateemethanol (v/v)): Rf¼0.28. Mp 144e145 ꢀC. H
NMR (300 MHz, DMSO-d6, d (ppm)): 8.22 (dd, J¼8.0,
1.6 Hz, 1H, H-9), 8.17 (dd, J¼8.0, 1.6 Hz, 1H, H-6), 8.10
(d, J¼10.0 Hz, 1H, H-1), 8.04 (dd, J¼10.0, 2.4 Hz, 1H, H-
2), 8.00 (d, J¼2.4 Hz, 1H, H-4), 7.85 (m, 2H, H-7, H-8),
3.64 and 3.60 (2s, 6H, N(CH3)2). 13C NMR (75 MHz,
DMSO-d6, d (ppm)): 156.1, 144.1, 139.8, 139.6, 138.0,
134.6, 133.2, 129.8, 126.3, 126.1, 125.8, 109.7, 43.4, 42.9.
IR (film, n (cmꢂ1)): 2800, 1617, 1559, 1489, 1429, 1411,
1252, 1118, 1078, 887, 835, 772. LRMS (ESI): m/z calcd
for C14H13N2S [M]þ 241.08, found 241.1. Anal. Calcd for
C14H13N2SI3: C, 27.03; H, 2.11; N, 4.50; S, 5.15; I, 61.20.
Found C, 27.12; H, 1.97; N, 4.46; S, 5.23; I, 60.94.
Distilled, deionized water was utilized in the preparation of
all buffers and all aqueous reactions. Chemicals were of the
highest available purity and were used without further purifica-
tion. Anhydrous magnesium sulfate, cesium carbonate, chloro-
form, a,a0-dibromo-p-xylene, dichloromethane, dimethylamine
(2 M solution in methanol), dimethylformamide, ethidium bro-
mide, iodine, methanol, 10H-phenothiazine, 1-piperazinecarb-
aldehyde, potassium hydroxide, potassium carbonate, sodium
phosphate dibasic, and sodium phosphate monobasic were
obtained from the Aldrich Chemical Company. The transforma-
tion of Escherichia coli competent cells (Stratagene, XL-blue)
with pUC19 plasmid DNA (Sigma) and growth of bacterial
cultures in LB broth were performed according to published
procedures.33 A Qiagen Plasmid Mega Kit was used to purify
the plasmid DNA from the bacterial cultures. Ultra PureÔ calf
thymus (CT) DNA (Invitrogen Lot no. 15633-019, 10 mg/mL,
average sizeꢄ2000 bp) was used without purification. The
concentration of CT DNA solutions was determined by UVe
visible spectrophotometry using the extinction coefficient
4.2.3. 1,10-[1,4-Phenylenebis(methylene)]bis(4-piperazine-
carbaldehyde) (3)
A
0.010 mol),
solution containing a,a0-dibromo-p-xylene (2.60 g,
1-piperazine
carboxaldehyde
(2.30 mL,
0.022 mol), and potassium carbonate (1.40 g, 0.010 mol) in
40 mL of methanol was refluxed under an open atmosphere
for 24 h. Reaction progress was monitored by silica gel TLC.
After the starting materials were consumed, the reaction was
concentrated to dryness under reduced pressure. The resulting
solid was then dissolved in 25 mL of water and extracted with
dichloromethane (50 mL) three times. The organic layer was
dried over anhydrous magnesium sulfate, filtered, and concen-
trated under reduced pressure to yield 2.47 g (75%) of white
solid 3. TLC (44:8:1 chloroformemethanole28% ammonium
1
hydroxide): Rf¼0.69. Mp¼140e142 ꢀC. H NMR (300 MHz,
3260¼12,824 Mꢂ1 (bp) cmꢂ1
.
DMSO-d6, d (ppm)): 7.97 (s, 2H, NCHO), 7.25 (s, 4H, Ph),
3.47 (s, 4H, CH2ePh), 3.38e3.30 (m, 8H, CH2-a), 2.34 and
2.28 (2 t, J¼4.9 Hz, 4H, CH2-b). 13C NMR (75 MHz, DMSO-
d6, d (ppm)): 160.9 (CHO), 136.7 (Cipso Ph), 128.9 (Ph), 61.8
(CH2ePh), 53.3 and 52.1 (CH2-a), 44.9 and 39.4 (CH2-b). IR
(film, n (cmꢂ1)): 2870, 2805, 1668, 1514, 1489, 1431, 1400,
1226, 1125, 1017, 998, 841, 821, 770. LRMS (ESI): m/z calcd
for C18H27N4O2 [MþH]þ 331.21, found 331.1.
4.2. Synthesis
4.2.1. Phenothiazin-5-ium tetraiodide hydrate (1)
A solution of 10H-phenothiazine (0.566 g, 2.84 mmol) in
20 mL of chloroform was stirred at 5 ꢀC and iodine (2.16 g,
8.51 mmol) dissolved in 50 mL of chloroform was added drop-
wise over 1 h. The reaction mixture was stirred at 5 ꢀC for an
additional hour, while its progress was monitored by silica gel
TLC. The resulting precipitate was then filtered, washed with
a copious amount of chloroform, and was dried overnight in
vacuo to afford 1.63 g (80%) of dark-blue solid 1. TLC (chloro-
form): Rf¼0.09. Mp¼170 ꢀC (dec). 1H NMR (300 MHz,
4.2.4. 1,10-[1,4-Phenylenebis(methylene)]bispiperazine (4)
A solution of 3 (0.500 g, 1.50 mmol) in 3 mL of ethanol
and 2 mL of water was treated with potassium hydroxide
(0.421 g, 7.50 mmol) and refluxed under an open atmosphere
for 24 h. Reaction progress was monitored by silica gel

B. Wilson et al. / Tetrahedron 64 (2008) 3429e3436
3435
TLC. After the starting material was consumed, the reaction
mixture was allowed to cool and was then extracted with di-
chloromethane (10 mL) three times. The organic layer was
dried over anhydrous magnesium sulfate, filtered, and concen-
trated under reduced pressure to yield 0.321 g (78%) of white
solid 4. TLC (44:8:1 chloroformemethanole28% ammonium
13C NMR (75 MHz, DMSO-d6, d (ppm)): 154.3 and 153.0 (C-
3, C-7), 138.2 and 138.1 (C-1, C-9), 135.9, 135.1, 134.2 and
133.9 (CipsoePh, C-4a, C-5a, C-9a and C-10a), 129.1 (Ph),
119.9 and 119.0 (C-2, C-8), 107.3 (C-4, C-6), 61.2 (CH2ePh),
52.5 (CH2-b), 47.5 (CH2-a), 41.5 (NCH3). IR (film, n (cmꢂ1)):
2954, 2884, 2834, 1590, 1488, 1385, 1347, 1228, 1140, 1035,
1001, 881, 778. HRMS (ESI): m/z calcd for C44H48N8S2
[M]2þ 376.1722, found 376.1725.
1
hydroxide): Rf¼0.45. Mp¼146e148 ꢀC. H NMR (300 MHz,
DMSO-d6, d (ppm)): 7.19 (s, 4H, Ph), 3.36 (s, 4H, CH2e
Ph), 2.64 (t, J¼4.6 Hz, 8H, CH2-a), 2.23 (t, J¼4.6 Hz, 8H,
CH2-b). 13C NMR (75 MHz, DMSO-d6, d (ppm)): 136.7 (Cipso
Ph), 128.6 (Ph), 62.6 (CH2ePh), 54.1 (CH2-b), 45.6 (CH2-a).
IR (film, n (cmꢂ1)): 3244, 2880, 2804, 1514, 1453, 1360,
1335, 1249, 1134, 1008, 858, 830, 765. LRMS (ESI): m/z
calcd C16H27N4 [MþH]þ 275.22, found 275.2.
4.3. UVevisible spectrophotometry
The extinction coefficients of compounds 5 and 6 were
obtained using 500 mL solutions containing 1e10 mM of dye
and 10 mM sodium phosphate buffer pH 7.0 in the absence
and presence of 38e380 mM bp CT DNA. The solutions
were pre-equilibrated for 12 h (rt, no hn), after which spectra
were recorded at 22 ꢀC in 1 cm quartz cuvettes and absorbance
was plotted as a function of concentration. Linear least square
fits to the data yielded slopes that were averaged over three tri-
als (KaleidaGraph version 3.6.4 software). Using the proce-
dure described above, extinction coefficients in the absence
of calf thymus DNA were also recorded in the presence of a
final concentration of aqueous 1% SDS (w/v).
4.2.5. 7-Dimethylamino-3-(1,10-[1,4-phenylenebis-
(methylene)bispiperazine])phenothiazin-5-ium iodide (5)
A solution of 2 (0.507 g, 0.815 mmol) and 4 (0.224 g,
0.815 mmol) in 25 mL of methanol was vigorously stirred
for 72 h at rt. Progress was monitored by silica gel TLC.
Upon completion, the reaction was concentrated under re-
duced pressure and the resulting solid was purified by flash
column chromatography (diameter 4 cm, 70 g silica gel) using
9.5:0.5 dichloromethaneemethanol as the eluent to afford
170 mg (34%) of dark-blue solid 5. TLC (9.5:0.5 dichlorome-
4.4. Viscometric measurements
1
thaneemethanol): Rf¼0.2. Mp>300 ꢀC. H NMR (300 MHz,
DMSO-d6, d (ppm)): 7.90 (m, 2H, H-1, H-9), 7.68 (br s, 1H,
H-6), 7.62 (dd, J¼9.6, 1.8 Hz, 1H, H-8), 7.53e7.49 (m, 2H,
H-2, H-4) 3.85 (br s, 4H, CH2-a), 3.54 (s, 4H, CH2ePh),
3.37 (s, 6H, NCH3), 3.09 (t, J¼5.0 Hz, 4H, CH2-a0), 2.55
(br, 8H, CH2-b,b0). 13C NMR (75 MHz, DMSO-d6,
d (ppm)): 154.2 and 153.0 (C-3, C-7), 138.1 and 138.0 (C-1,
C-9), 136.7, 135.8, 135.0 and 134.1 (CipsoePh, C-4a, C-5a,
C-9a, and C-10a), 129.0 (Ph), 119.7 (C-2), 118.9 (C-8),
107.2 and 107.1 (C-4, C-6), 61.2 and 61.1 (CH2ePh), 52.4
(CH2-b0), 49.1 (CH2-b), 47.4 (CH2-a), 43.2 (CH2-a0), 41.4
(NCH3). IR (film, n (cmꢂ1)): 1743, 1593, 1487, 1386, 1352,
1229, 1136, 1042, 990, 882, 825, 779. HRMS (ESI): m/z calcd
for C30H38N6S [MþH]2þ 257.1439, found 257.1462.
In a total volume of 1 mL, individual solutions containing
200 mM bp of CT DNA (average lengthꢄ2000 bp) and
10 mM sodium phosphate buffer pH 7.0 in the absence and
presence of 2e12 mM of compound 5 or 6 were allowed to
pre-equilibrate for 12 h (rt, no hn). A Cannon-Ubbelohde
size 75 capillary viscometer immersed in a thermostated water
bath set at 25ꢅ0.2 ꢀC was then used to measure DNA viscos-
ity. The flow times of the buffer, DNA in buffer, and dyee
DNA in buffer were measured with a stopwatch and were
averaged over four trials to an accuracy of ꢅ0.2 s. After sub-
tracting the averaged flow time of the buffer, DNA (h0) and
dyeeDNA (h), averaged flow times were plotted as (h/h0)1/3
versus the molar ratio of dye to DNA bp. Slopes were obtained
by conducting linear least square fits to the data (Kaleida-
Graph version 3.6.4 software). The conventional method for
performing viscosity assays involves the gradual titration of li-
gand into a DNA solution inside a single viscometer. In the
procedure reported here, a series of individual DNA solutions
containing different concentrations of ligand were prepared
and allowed to pre-equilibrate before the viscosity of each
solution was measured. This alternative technique may be par-
ticularly useful for compound 6 as well as for other phenothi-
azines (e.g., 1,9-dimethyl methylene blue and methylene blue)
for which pre-equilibration with DNA is required to disrupt
ligand dimerization.4,6,22,23
4.2.6. N,N0-Bis[(7-dimethylamino)phenothiazin-5-ium-3-
yl]-1,10-[1,4-phenylenebis(methylene)bispiperazine]
diiodide (6)
To a solution of 2 (0.200 g, 0.322 mmol) in 20 mL of dime-
thylformamide were added 4 (0.044 g, 0.160 mmol) and
cesium carbonate (0.156 g, 0.480 mmol). The reaction was vig-
orously stirred at rt for 48 h and then concentrated under reduced
pressure. The progress of the reaction was monitored by silica
gel TLC. The resultant solid was purified via flash column chro-
matography (diameter 3 cm, 39 g silica gel) using 9.5:0.5 di-
chloromethaneemethanol as the eluent to afford 66 mg (41%)
of dark-blue solid 6. TLC (9.5:0.5 dichloromethaneemethanol):
Rf¼0.3. Mp>300 ꢀC. 1H NMR (300 MHz, DMSO-d6, d (ppm)):
7.93 (app d, J¼9.7 Hz. 4H, H-1, H-9), 7.70 (d, J¼2.7 Hz, 2H,
H-6), 7.63 (dd, J¼9.7, 2.7 Hz, 2H, H-8), 7.55e7.51 (m, 4H,
H-2, H-4), 7.32 (s, 4H, Ph), 3.85 (br s, 8H, CH2-a), 3.56 (s,
4H, CH2ePh), 3.38 (s, 12H, NCH3), 2.57 (br s, 8H, CH2-b).
4.5. DNA photocleavage
Individual solutions containing 38 mM bp of pUC19 plasmid
DNA, 10, 5, 2, 1, 0.5, 0.25, and 0.0 mM concentrations of com-
pound 5 or 6, and 10 mM sodium phosphate buffer pH 7.0 were

3436
B. Wilson et al. / Tetrahedron 64 (2008) 3429e3436
´
3. Tuite, E.; Norden, B. J. Am. Chem. Soc. 1994, 116, 7548.
prepared in a total volume 20 mL. After a 12 h pre-equilibration
period (rt, no hn), the samples were kept in the dark or aerobi-
cally irradiated for 60 min in a ventilated Rayonet Photochem-
ical reactor fitted with twelve 575 nm lamps (spectral output
400e650 nm; Southern New England Ultraviolet Co.). In
time course experiments, individual, pre-equilibrated 20 mL so-
lutions consisting of 10 mM of 5 or of 6 and 38 mM bp pUC19
plasmid DNA in 10 mM sodium phosphate buffer pH 7.0
were aerobically irradiated as described above for 0, 5, 10,
20, 30, 40, 50, and 60 min. Cleaved DNA products were elec-
trophoresed on 1% nondenaturing agarose gels containing
ethidium bromide (0.5 mg/mL), visualized on a transilluminator
set at 302 nm, photographed, and scanned. ImageQuant version
5.2 software (Amersham Biosciences) was then used to quanti-
tate supercoiled and nicked plasmid DNA. The density of the
supercoiled DNA was multiplied by a correction factor of
1.22 to compensate for the decreased binding affinity of ethid-
ium bromide to supercoiled versus nicked DNA forms. Photo-
cleavage yields were calculated from the data according to
the formula [nicked DNA/total DNA]ꢁ100.
4. Tuite, E. M.; Kelly, J. M. J. Photochem. Photobiol., B: Biol. 1993, 21, 103.
5. Harris, F.; Chatfield, L. K.; Phoenix, D. A. Curr. Drug Targets 2005, 6,
615.
6. Tardivo, J. P.; Del Giglio, A.; Oliveira, C. S.; Gabrielli, D. S.; Junqueira,
H. C.; Tada, D. B.; Severino, D.; Turchiello, R. F.; Baptista, M. S. Photo-
diagn. Photodyn. Ther. 2005, 2, 175.
7. OhUigin, C.; McConnell, D. J.; Kelly, J. M.; van der Putten, W. J. M.
Nucleic Acids Res. 1987, 15, 7411.
8. Devasagayam, T. P.; Steenken, S.; Obendorf, M. S.; Schulz, W. A.; Sies,
H. Biochemistry 1991, 30, 6283.
9. Juzeniene, J.; Moan, A. Photodiagn. Photodyn. Ther. 2007, 4, 3.
10. Tardivo, J. P.; Del Giglio, A.; Paschoal, L. H.; Baptista, M. S. Photomed.
Laser Surg. 2006, 24, 528.
11. Jacob, H. Photochem. Photobiol. 1971, 14, 743.
12. Abe, H.; Wagner, S. J. Photochem. Photobiol. 1995, 61, 402.
13. Wainwright, M.; Mohr, H.; Walker, W. H. J. Photochem. Photobiol., B
2007, 86, 45.
14. Hall, D. B.; Kelley, S. O.; Barton, J. K. Biochemistry 1998, 37, 15933.
15. Portugal, J.; Cashman, D. J.; Trent, J. O.; Ferrer-Miralles, N.; Przewloka,
T.; Fokt, I.; Priebe, W.; Chaires, J. B. J. Med. Chem. 2005, 48, 8209.
16. Carrondo, M. A.; Coll, M.; Aymami, J.; Wang, A. H.; van der Marel,
G. A.; van Boom, J. H.; Rich, A. Biochemistry 1989, 28, 7849.
17. Wilson, W. D.; Barton, H. J.; Tanious, F. A.; Kong, S. B.; Strekowski, L.
Biophys. Chem. 1990, 35, 227.
18. Strekowski, L.; Hou, D.-F.; Wydra, R. L. J. Heterocycl. Chem. 1993, 30,
1693.
4.6. Thermal melting studies
19. McConnaughie, A. W.; Spychala, J.; Zhao, M.; Boykin, D.; Wilson, W. D.
J. Med. Chem. 1994, 37, 1063.
Individual 3 mL solutions containing 10 mM sodium phos-
phate buffer pH 7.0 and 12.5 mM bp calf thymus DNA in the ab-
sence and presence of 7.50 mM of compound 5 or 6 were placed
in 3 mL (1 cm) quartz cuvettes obtained from Starna. After
a 12 h pre-equilibration period (rt, no hn), absorbance was mon-
itored at 260 nm while the DNAwas melted using a Peltier heat
block to increase the temperature from 25 to 100 ꢀC at a rate of
0.5 ꢀC minꢂ1. KaleidaGraph version 3.6.4 software was then
used to calculate the first derivative of a plot of DA260/DT versus
temperature. The maximum of the first derivative plot was used
to identify the Tm value of the DNA.
20. Leventis, N.; Chen, M.; Sotiriou-Leventis, C. Tetrahedron 1997, 53,
10083.
21. Mellish, K. J.; Cox, R. D.; Vernon, D. I.; Griffiths, J.; Brown, S. B. Photo-
chem. Photobiol. 2002, 75, 392.
22. Jockusch, S.; Lee, D.; Turro, N. J.; Leonard, E. F. Proc. Natl. Acad. Sci.
U.S.A. 1996, 93, 7446.
23. Mohammad, T.; Morrison, H. Bioorg. Med. Chem. Lett. 1999, 9, 2249.
24. Suh, D.; Chaires, J. B. Bioorg. Med. Chem. 1995, 3, 723.
25. Cohen, G.; Eisenberg, H. Biopolymers 1969, 8, 45.
26. Crothers, D. M. Biopolymers 1968, 6, 575.
27. Wilson, W. D. Nucleic Acids in Chemistry and Biology; Blackburn,
G. M., Gait, M. J., Eds.; Oxford University Press: New York, NY,
1996; pp 331e370.
˚
28. A linker length of 11 A was obtained by measuring the distance in
Acknowledgements
between the two piperazine ring N-1 atoms in a three-dimensional
structure of compound 6 (CambridgeSoft Corporation Chem 3D Pro
software version 5.0).
This work was supported by the NSF (CHE-9984772;
K.B.G.), the CICYT (Project BQU 2002-02576; A.L.) and
the U.S. Department of Education (GAANN Fellowship;
B.W.). We are grateful for assistance from Professors Markus
W. Germann, Thomas L. Netzel, Lucjan Strekowski, W. David
Wilson (G.S.U.) and from Dr. Mijail V. Galakhov (U.A.H.).
29. Wilson, W. D. Comprehensive Natural Products Chemistry: DNA and
Aspects of Molecular Biology; Barton, D. H. R., Meth-Cohn, O., Naka-
nishi, K., Kool, E. T., Eds.; Elsevier Science: Oxford, 1999; Vol. 7,
pp 427e476.
´
30. Wilson, W. D.; Tanious, F. A.; Fernandez-Saiz, M. DrugeDNA Interaction
Protocols; Fox, K. R., Ed.; Humana: New Jersey, NJ, 1997; pp 219e240.
31. Wilson, W. D.; Ratmeyer, L.; Zhao, M.; Strekowski, L.; Boykin, D. Bio-
chemistry 1993, 32, 4098.
References and notes
32. Harris, R. G.; Wells, J. D.; Johnson, B. B. Colloids Surf. A: Physiochem.
Eng. Asp. 2001, 180, 131.
33. Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning: A Labora-
tory Manual, 2nd ed.; Cold Spring Harbor: New York, NY, 1989.
1. Weber, J. M.; Rawls, M. T.; MacKenzie, V. J.; Limoges, B. R.; Elliott,
C. M. J. Am. Chem. Soc. 2007, 129, 313.
2. Genwa, K. R.; Chouhan, A. J. Chem. Sci. 2004, 116, 339.